Comfort Systems USA Is a $1,600 Stock Almost Nobody Owns. Here's Why It Keeps Crushing the Market.

Comfort Systems USA's backlog reached $14.06 billion, exceeding Wall Street's forecast of just under $13 billion in 2026 revenue and supporting near-term growth visibility. Technology-related revenue, which includes data-center work, rose to 58% of revenue from 40% a year earlier, underscoring its exposure to AI infrastructure spending. The company’s reported backlog excludes recurring service and maintenance contracts, implying additional long-term revenue potential, although installation-project backlog typically converts within less than one year.

Analysis

FIX’s valuation now embeds continued high-teens-to-20% growth well beyond the current project cycle; the relevant variable is not reported backlog coverage but quarterly book-to-bill and the mix of large, fixed-price technology projects. The company’s prefabrication capability should preserve labor productivity and margins versus smaller mechanical contractors, but accelerated hyperscaler schedules can create working-capital drag, subcontractor cost inflation, and execution risk before any revenue slowdown appears. Near term, the shares are vulnerable to even a modest deceleration in orders because the backlog converts quickly and offers limited duration protection.

The underappreciated structural upside is the installed-base annuity: commissioned AI facilities require recurring mechanical, electrical, cooling, controls, retrofit, and capacity-upgrade work that is less visible than project awards. That supports a higher through-cycle margin and multiple only if service revenue becomes material enough to offset the eventual normalization of new-build activity. Likely read-through beneficiaries include EMCOR (EME) and Quanta Services (PWR), while less-specialized private and public mechanical peers face labor and procurement pressure without FIX’s scale or prefabrication advantage.

Consensus appears to treat every hyperscaler capex announcement as fungible demand for FIX. The key distinction is whether spending is directed toward power availability and utility interconnection rather than facilities construction; grid bottlenecks could defer building starts despite headline capex remaining elevated. Over 1-3 months, monitor technology/industrial bookings, gross-margin guidance, receivables and contract-asset growth, and customer concentration; over 6-18 months, service attach rates and data-center retrofit awards determine whether the premium is durable.

Key Decisions for Investors

• Do not chase FIX after a parabolic run; establish only on a 10-15% pullback or after a quarterly technology book-to-bill above 1.1x with stable gross-margin guidance. Target a 6-12 month long only if order growth validates the multiple; exit on two consecutive quarters below 1.0x book-to-bill or a meaningful margin-guide cut.
• Use a relative-value basket rather than outright beta: long EME / short FIX in equal dollar amounts over the next 1-3 months if FIX continues to outperform without a corresponding order acceleration. EME offers similar mission-critical electrical exposure at potentially lower single-name AI-project concentration; cover if FIX demonstrates sustained service-margin expansion or EME misses on data-center bookings.
• For AI-infrastructure exposure, prefer staged long PWR over incremental FIX at current levels for a 6-18 month horizon. Power interconnection and transmission constraints are a prerequisite for data-center energization and may capture spend even if facility construction timing slips; thesis fails if regulated-utility capex plans or PWR’s backlog guidance weaken.
• Set an earnings alert around contract assets/receivables growth exceeding revenue growth by more than 10 percentage points. That would signal cash-conversion and fixed-price execution risk, warranting reducing FIX exposure even if reported revenue and backlog remain strong.